ECU choice

[Electro]

Greetings:
I am trying to choose an ECU for this coming year's team. There's a plethora of choices and all kinds of options amongst them.
My questions per topic:

Fuel, timing maps:
what kind of resolution do other teams run? I've seen as little as 16x16 maps but others have far greater. It seems obvious that more resolution is better for smoothing the map and maximizing performance; just looking for what everyone else has seen. What methods do you use for tuning? Dynomometers are great, but what about for drivability? How does one go about tuning after saying "hey, when I barely touch the gas out of a turn at 8k, I lose my rear end".


Fundamental options:
Dwell time: Which ECU's have adjustment, which don't? Which ECU is going to screw me over at high RPMS because the dwell's way off

Closed Loop: Closed loop on a race engine; great for tuning but ultimately don't we just tune for best power?, how big is the environmental variable? Wideband or narrow? Standalone for tuning, eliminate for racing?

Map Variables:
MAF or Pressure? sure, MAF is awesome, so why wouldn't all ECU;s offer it? We're not running turbo so it shouldnt be an issue.

Piggyback versus standalone:
Is anyone running the powercommander or similar? It seems really useful, as the stock ECU has already taken care of the startup mapping, the throttle off cut, etc. The extra functions are great on a standalone but I can implement them seperately.


We've used the AEM in past years but we're trying to go cheapers and eliminate all the functions we don't need (of which there are a ton) Just wanted to see what everyone thought of their ECU's.
Those I've looked into:
Haltech
AEM
Performance electronics
powercommander (piggyback)
MOTEC

Thanks guys

[Electro]

Greetings:
I am trying to choose an ECU for this coming year's team. There's a plethora of choices and all kinds of options amongst them.
My questions per topic:

Fuel, timing maps:
what kind of resolution do other teams run? I've seen as little as 16x16 maps but others have far greater. It seems obvious that more resolution is better for smoothing the map and maximizing performance; just looking for what everyone else has seen. What methods do you use for tuning? Dynomometers are great, but what about for drivability? How does one go about tuning after saying "hey, when I barely touch the gas out of a turn at 8k, I lose my rear end".


Fundamental options:
Dwell time: Which ECU's have adjustment, which don't? Which ECU is going to screw me over at high RPMS because the dwell's way off

Closed Loop: Closed loop on a race engine; great for tuning but ultimately don't we just tune for best power?, how big is the environmental variable? Wideband or narrow? Standalone for tuning, eliminate for racing?

Map Variables:
MAF or Pressure? sure, MAF is awesome, so why wouldn't all ECU;s offer it? We're not running turbo so it shouldnt be an issue.

Piggyback versus standalone:
Is anyone running the powercommander or similar? It seems really useful, as the stock ECU has already taken care of the startup mapping, the throttle off cut, etc. The extra functions are great on a standalone but I can implement them seperately.


We've used the AEM in past years but we're trying to go cheapers and eliminate all the functions we don't need (of which there are a ton) Just wanted to see what everyone thought of their ECU's.
Those I've looked into:
Haltech
AEM
Performance electronics
powercommander (piggyback)
MOTEC

Thanks guys

[BStoney]

Performance Electronics or MOTEC are going to be your most common out there, but I would say definitely PE if you want the best bang for the buck and ease of use as well as tech support.

...also to your comment about tuning for best power, I feel you are way off here. This competition is about torque, not horsepower. Therefore, you should be tuning for max torque at each spot in your map.

My .02...

[Jonathan R.]

For the Motec M4 the fuel and ignition map size is 20 x 27.

Most teams use a dynamometer to tune their engine.

With the M4 you can adjust the dwell time with a table if you want with the Motec, but usually it's something you don't really play around with.

Since it's the engine performance that interest you, there is no point running in close loop.
As for the O2 sensor, a narrow-band can only sense air/fuel ratio over a small range (about 14.5 to 15.0). On the other hand, Wide-bands go from about 10.3 to infinite. So for tuning, a wide-band is really worth it.

You can have both MAF and a pressure sensor.

Datalogging is also a feature you should consider in your ECU choice.

[VFR750R]

To clarify on the max power. Tuning for max torque at at every point is also tuning for max power at every point.

[Electro]

Agreed on the power comment. Tuning for best torque across the board is tuning for best power. Merely semantics regarding torque at RPM versus just torque.
My comment on Dwell Time was due to an experience with our AEM system being set improperly and RPM limiting our engine due to saturating the coils for too long.
AEM had some nice logging features, MOTEC's is of course great but too expensive... Any others with simple 02, throttle position, and few other logged parameters? I don't really care about auxillary inputs

Why would you want both MAF and pressure? MAF being the more reliable of the two options, why not just stick with it on all ECU's (shy of turbos)?

Thanks

[Jonathan R.]

And why would you like a MAF? It's to get some data, same with the pressure sensor.

EDIT : Ok I misunderstood, you're running your engine with a MAF, I thought it was just for acquisition and the engine was running with a MAP.

[Sam Zimmerman]

Buy the PE ECU.

Screw resolution.

Tell your driver to quit whining and learn to drive.

Dwell time is not important.

Closed loop all the way.

Narrow band.

Tune with MAF.

Stand alone.

Or...figure out what your goals are (drivability, simplicity, power, proving someone wrong), gather information, and use the best decision making process available to you to make the decision. Don't try to solve all your problems at once, your reaction to your driver can wait.

Use the forum to ask questions like, "I feel a smooth, responsive map is a higher priority than trying to bleed the last bit of torque (or power) out of every single point because...what is your opinion?" or "I am just learning, which ECS has a reputation for good tech support?" or my favorite, "Here is my theory based on this research, where do you disagree." If you ask questions like, "which ECS should I use?" you will get responses based on other people's goals, not yours. How valuable is that?

Just a friendly piece of advice for a newbie, take it or leave it. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[BeaverGuy]

Personally, I would want MAF in every situation as it is a direct measurement but it is a more expensive and delicate sensor. Additionally only a handful of ECUs offer it. Motec M8 or M800 are the only ones I can think of at the moment, autronic might and I imagine Bosch does. By Looking at the Haltech software they might support it too but I can't be certain. I have read that GM uses a combined Speed Density and MAF setup on some of their vehicles but the reasoning for it was somewhat fuzzy.

Dwell for me was a set it and forget it type of thing. The only time it should come into issue is if you have it set too high or are running wasted spark and running really high RPMs. I calculated the point where you wouldn't be able to fully charge the stock COP units at one point in time and I seem to remember it being above the usable RPM range of a 600.

I guess I'll give the reasons that I picked the ECU I did for the '05 car.

#load & RPM points: We picked a 32X32 because it allowed us to have fairly precise control over the RPM range we wanted. We also wanted to be able to adjust the position of RPM points in order to fine tune troubled areas. We didn't end up using this feature primarily because of time constraints but it would have been beneficial had we had the time.

Injector outputs/Fueling Scenarios: we wanted something capable of sequential injection as our injector times with E85 and semi-sequential were at the edge of acceptable for a peak and hold injector at low loads and RPMs and switching to gas would further reduced our injector times. AS it was at low loads and RPMs we were still close to the point where the injector is simply opening and closing with no time at fully open.

Crank Sensor Type: Due to the location and design of our pre-existing dry sump coupling we had pretty much locked ourselves into a certain type of crank sensor so the ECU had to be able to support it. Not an issue for most teams but some ECUs are restricted to a single crank position sensing scheme and it may not be what the stock motor has.

Inputs and outputs and datalogging:Having this ability in the ECU means reduced reliance on another system and it can be useful for a tach, shift light, automatic shift cuts, fans, electric water pumps. We selected one with far more I/O than what we thought we needed. At the time it was really one of the few that had enough I/O to meet our needs but at one point we were trying to figure out if it met all of our I/O needs as it was it was slightly defficient in our desire. If it had another digital input it could replace our data aq system completely. Also, last years car will hopefully become a testbed for the future so having more I/O than you need for this year will help when you add that fancy electric whatachamahoozit to the car for testing.

For the wide band a built in controller wasn't really even considered it would have been a nice plus but not necessary. As it was we used a stand alone unit for tuning the engine on the dyno and did log some data with it connected to a spare input on the ECU. We didn't really consider logging in car O2 but we could have if we wanted or thought to do so and I would suggest doing it to verify transients.

Tuning for drivability can be done on the dyno. We did this a little bit by changing intake geometry. A large jump or dip in the torque curve will corespond to a point where the driver says the car is too sensitive or not sensitive enough to throttle inputs. You can tune this out via fuel before the engine ever runs in the car if you want. Other ways to deal with drivability issues related to throttle control include changing the throttle linkage or design.

The only tuning we did of the engine once it was in the car was for idle and startup characteristics. Though our dyno room flowed in air from outside the setup was different enough that our perfect start and idle on the dyno didn't translate to the car.

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by BeaverGuy:
#load & RPM points: We picked a 32X32 because it allowed us to have fairly precise control over the RPM range we wanted. We also wanted to be able to adjust the position of RPM points in order to fine tune troubled areas. </div></BLOCKQUOTE>

I have to ask, did you really visit 1024 load/ rpm sites on your dyno? Don't dismiss an ECU just because it 'only' has 12x12 or 16x16 base maps. As long as 1) you can define your rpm and load breakpoints and 2) the ECU interpolates between sites (and most do both) you will be fine.

In most cases there will be large areas of the base maps that are uninteresting. Save your dyno tuning investment for the 'interesting' areas (idle, intake/exhaust resonances, 'on cam' etc).

Datalogging capability is very important for track (read transient) tuning. A wideband lambda is not required for control (and if it is offered for control be very wary and have a bullet-proof installation, and read upon 'transport delay' in all your control textbooks...) but is very important for datalogging based tuning. So, if your ECU has an analogue input you can just buy a stand-alone wideband and log it that way.

Look carefully at the transient tuning options any ECU offers you. You need to be able to fine-tune acceleration/deceleration fuelling to compensate for manifold filling and fuel puddling (wall wetting) effects. Unless you have a very nice (read expensive) transient dyno you will have to do this on-track with your datalogging.

Dwell options can be confusing. Coil charging is time based not crankangle based. Yet the old points system set dwell based on a crank angle at the distributor. So you got too much dwell at low rpm and not enough at high rpm. Most ECUs (you should check) will specify dwell in ms so you just spend 10mins with an oscilloscope checking your coil charge and you are set with one dwell value for your entire rpm range. The only caveat is if your coils take so long to charge they can't fully charge in one engine cycle at max rpm (for COP, 1/2 a cycle for wasted spark). In that case you need different coils, not a different ECU.

Hope this all helps,
Regards, Ian

[BeaverGuy]

We didn't hit all the load points on the dyno and that wasn't as important as the RPM points. But stepped through every RPM point from 2000-12,500 which was 22 points and we adjusted below 2000 for idle, this year they tuned up to 13,500 at every 500 RPM.

[Poe]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I would want MAF in every situation as it is a direct measurement but it is a more expensive and delicate sensor. Additionally only a handful of ECUs offer it. Motec M8 or M800 are the only ones I can think of at the moment, autronic might and I imagine Bosch does. </div></BLOCKQUOTE>

We are running MAF on our M4

[BStoney]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by VFR750R:
To clarify on the max power. Tuning for max torque at at every point is also tuning for max power at every point. </div></BLOCKQUOTE>

Exactly why I am not an engine guy!! http://fsae.com/groupee_common/emoticons/icon_eek.gif

But, is it not easier to measure torque output on an engine dyno? Sure you can get the power from the torque since it's just a calculation...

[VFR750R]

Whichever, it's just a technicality http://fsae.com/groupee_common/emoticons/icon_wink.gif The point i was trying to make was torque is often mislabeled a low to midrange rpm trait, which is one of my petpeeves I guess. AVG HP (over your operating range) is a great goal, and although avg torque is similar you'll find with the RPM multiplier aren't the same. The best especially in a non-drag situation like ours is area under the torque curve.

[Electro]

Zimmerman,
Thanks for the input (I'm really just getting a feel for everyone else's goals while making my own).
Narrow band: If I tune with narrowband, do I just get close by common sense and use narrowband to fine tune? Doesn't that take a while? Standalone or ECU function?

Tune with MAF: I didn't think the PE had MAF capability.

Closed loop: Why such certainty? The only reason I see it necessary is for environmental factors, but if we have an intake air temp scaler, what's it matter? Are you saying closed loop just lets us get close with the tune and it'll make it better on the fly?

Appreciate the advice guys

[Electro]

Regarding resolution:
Interpolation between say 16x16 maps of values. Anyone had any experience with an increase of performance and consistency by just improving resolution (upgrading ecu).

[Eshu]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by BStoney:
But, is it not easier to measure torque output on an engine dyno? Sure you can get the power from the torque since it's just a calculation... </div></BLOCKQUOTE>

Can anybody think of a way to measure power directly on a dyno? Bueller? Bueller?

Electro, you might want to get a hold of Brian Lewis at Performance Electronics.

Out

[LCP_KL]

i guess we are one of the few who still run the primitive PC3 with ONLY fuel compensation. why? cos our whole engine team is made up of only 3 persons, and half of the time, we are machining our own components...

and guess what? we run it on a homemade dyno, and holding rpms mean balancing the load valve on the fly. makes for an exciting evening/weekend. haha.

we did invest in the Tech Edge wideband, with a useful 3+1 channel input.

although our fuel map resolution was fairly primitive, compared to the 32x32 maps around, it did proved more than adequate and together with many many on-the-road transient datalogged runs and tweaks, the engine ran very well.

and best thing of all is, the engine NEVER failed to start on the first try. NEVER.

[BeaverGuy]

I have a hard time believing Closed Loop control will help any time other than when you are idling around with low engine speed and throttle position. Widebands and their controllers would have to be incredibly fast and predictive to compensate for the situations encountered during any of the dynamic events.

An O2 sensor of any type isn't a requiremnt for tuning your engine if you are on a dyno. However, it is always nice to know what is going on.

I would agree that an engine team with only 3 people is rough. For my senior project I was one of 2 Senior MEs and we had help from another senior in Industrial engineering there was a lot to do. I wouldn't want to do it again but if I had the chance I would. I can't imagine trying to tune the engine on a dyno with a manual load valve though.

Anyway I thought I'd post the Excel spreadsheet comparing the ECUs we looked at when we selected ours in '05.ECU spreadsheet (http://home.comcast.net/%7Ejoshuagillett/ECU_features.xls)

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Eshu:
Can anybody think of a way to measure power directly on a dyno? </div></BLOCKQUOTE>

How about a trailer dyno? P=FV as well as P=nT...

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by BeaverGuy:
I have a hard time believing Closed Loop control will help any time other than when you are idling around with low engine speed and throttle position. Widebands and their controllers would have to be incredibly fast and predictive to compensate for the situations encountered during any of the dynamic events. </div></BLOCKQUOTE>

1) Road car ECUs operate in closed-loop with a narrow-band sensor only under specific circumstances, generally low-transient situations like idle, cruise, part-throttle acceleration. WOT is usually open-loop rich for peak power, also high speed cruise can be open-loop rich to cool the catalyst.

2) Motorsport/aftermarket ECUs offering a wide-band closed-loop feature tend to work more like an 'adaptive' system rather than true closed-loop feedback control. Otherwise transient performance is very poor. Or, they offer a dyno only mode designed for steady-state conditions only.

Note that no oxygen sensor can distinguish between lean combustion, a misfire or an air leak into the exhaust. Consider very carefully the robustness of your sensor installation before committing to any form of closed-loop control.

Regards, Ian

[awhittle]

I have been running a MegaSquirt on a jetski for about year. I have the choice of wireless data logging up to 3/4 mile or to a Palm Vx. It runs MAF and has never had a Squirt related failure. See www.ncs-stl.com/xp (http://www.ncs-stl.com/xp) for pics.

AW

[Sam Zimmerman]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Electro:
Zimmerman,
Thanks for the input (I'm really just getting a feel for everyone else's goals while making my own).
Narrow band: If I tune with narrowband, do I just get close by common sense and use narrowband to fine tune? Doesn't that take a while? Standalone or ECU function?

Tune with MAF: I didn't think the PE had MAF capability.

Closed loop: Why such certainty? The only reason I see it necessary is for environmental factors, but if we have an intake air temp scaler, what's it matter? Are you saying closed loop just lets us get close with the tune and it'll make it better on the fly?

Appreciate the advice guys </div></BLOCKQUOTE>

Sorry, Electro. I thought it was clear that the one-liners weren't serious. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif Unless we get into a rousing discussion of intake design I rarely answer direct questions such as MAF or pressure and closed or open loop type of questions, mostly because there are plenty of opinions out there and there are as many right answers as there are options. What are your options, how do you rank them, what are your goals, etc.?

I will second the notion that you should call Brian Lewis @ PE. You will not find better customer service and I honestly think that if you give him an honest assessment of what you are looking for and why he will give you an honest assessment of where his equipment fits your needs and where it doesn't. Brian and his crew are as good as they come.

[Mike Cook]

Electro doesn't understand jokes. http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Electro]

This is true... it's a minor form of retardation.
They're looking for a cure. Every option involves social interaction though, not sure if I can handle it